# Difference between revisions of "OptSolution"

(→Examples) |
|||

Line 2: | Line 2: | ||

[[Category:Doc Status D]] <!-- For Lumina use, do not change --> | [[Category:Doc Status D]] <!-- For Lumina use, do not change --> | ||

− | |||

− | |||

− | |||

+ | == OptSolution(opt'', decision'') == | ||

Returns the solution to an optimization problem «opt» specified by [[DefineOptimization]]. | Returns the solution to an optimization problem «opt» specified by [[DefineOptimization]]. | ||

− | |||

− | Evaluating a variable that uses OptSolution() will trigger an attempt to solve the optimization problem, unless it has already been solved by another call to OptSolution, or a related function, such as [[OptStatusText]]. | + | If you specify as «decision» the name of a global Decision variable used in the Optimization, it returns the solution (optimal value) for that decision. If you omit «decision», it returns the solution to all Decision variables as a vector indexed by a local index '''.Vars''', which contains all the decision values flattened so that each solution is a scalar. |

+ | |||

+ | Evaluating a variable that uses [[OptSolution]]() will trigger an attempt to solve the optimization problem, unless it has already been solved by another call to [[OptSolution]], or a related function, such as [[OptStatusText]]. | ||

+ | |||

+ | [[OptSolution]] returns a result when it finds an optimal, or likely optimal solution. If it finds no feasible solution it gives a warning if you have [[Preferences|Show Result Warnings]] turned on. It is entirely possible that there is no solution, or that the solver could not find a feasible solution, in which cases the values returned by [[OptSolution]] are arbitrary. So, you should always check [[OptStatusText]] or [[OptStatusNum]] to check that it has found a feasible and optimal solution. | ||

− | + | If the Optimization used a local variable, say <code>D1</code>, as a Decision, declared as local in the expression using [[DefineOptimization]], for example | |

+ | :<code>Variable OptimizeIt := VAR d1 := 0; DefineOptimization(Decisions: d1; ....)</code> | ||

− | + | you can get its solution by defining a local with the same identifier, <code>D1</code>, preceding [[OptSolution]](), and giving <code>D1</code> as «decision»: | |

− | + | :<code>Variable D1_solution := VAR d1 := 0; OptSolution(OptimizeIt, d1)</code> | |

− | you can get its solution by defining a local with the same identifier, D1, preceding OptSolution(), and giving D1 as «decision»: | + | |

− | + | ||

− | = | + | == Example == |

+ | Find the minimum of the [[GammaFn]](x) for <code>x > 0</code>: | ||

− | + | :<code>Var x := 1;</code> | |

+ | :<code>Var opt := DefineOptimization(decisions: x, minimize: GammaFn(x), domain: Continuous(lb: 0));</code> | ||

+ | :<code>OptSolution(opt, x)</code> | ||

− | + | ==History== | |

− | + | This function was introduced in [[Analytica 4.3]], in earlier versions, use [[LpSolution]]. | |

− | + | ||

− | = See Also = | + | == See Also == |

− | * [[ | + | * [[DefineOptimization]] |

− | * | + | * [[OptStatusText]] |

− | * [[OptStatusText]] | + | * [[OptStatusNum]] |

+ | * [[OptStatusText]] | ||

+ | * [[OptStatusNum]] | ||

* [[OptObjective]] | * [[OptObjective]] |

## Revision as of 23:46, 20 January 2016

## OptSolution(opt*, decision*)

Returns the solution to an optimization problem «opt» specified by DefineOptimization.

If you specify as «decision» the name of a global Decision variable used in the Optimization, it returns the solution (optimal value) for that decision. If you omit «decision», it returns the solution to all Decision variables as a vector indexed by a local index **.Vars**, which contains all the decision values flattened so that each solution is a scalar.

Evaluating a variable that uses **OptSolution**() will trigger an attempt to solve the optimization problem, unless it has already been solved by another call to **OptSolution**, or a related function, such as OptStatusText.

**OptSolution** returns a result when it finds an optimal, or likely optimal solution. If it finds no feasible solution it gives a warning if you have Show Result Warnings turned on. It is entirely possible that there is no solution, or that the solver could not find a feasible solution, in which cases the values returned by **OptSolution** are arbitrary. So, you should always check OptStatusText or OptStatusNum to check that it has found a feasible and optimal solution.

If the Optimization used a local variable, say `D1`

, as a Decision, declared as local in the expression using DefineOptimization, for example

`Variable OptimizeIt := VAR d1 := 0; DefineOptimization(Decisions: d1; ....)`

you can get its solution by defining a local with the same identifier, `D1`

, preceding **OptSolution**(), and giving `D1`

as «decision»:

`Variable D1_solution := VAR d1 := 0; OptSolution(OptimizeIt, d1)`

## Example

Find the minimum of the GammaFn(x) for `x > 0`

:

`Var x := 1;`

`Var opt := DefineOptimization(decisions: x, minimize: GammaFn(x), domain: Continuous(lb: 0));`

`OptSolution(opt, x)`

## History

This function was introduced in Analytica 4.3, in earlier versions, use LpSolution.

Enable comment auto-refresher